What Is Plateau Pressure in Mechanical Ventilation?

Plateau pressure is the pressure measured in the airways at the end of a breath delivered by a mechanical ventilator, taken during a brief pause when no air is flowing. It reflects how much the lungs and chest wall are being stretched by the volume of air sitting inside them, stripped of the pressure needed just to push air through the tubing and airways. In critical care, it has become one of the main guardrails clinicians use to avoid damaging the lungs with the very machine keeping a patient alive. Yet plateau pressure is a rougher proxy for actual lung stress than many people realize, and a growing body of evidence suggests that reading it in isolation can mislead treatment decisions in important ways.

What Plateau Pressure Actually Tells You

During mechanical ventilation, the ventilator pushes a set volume of air into the lungs. While that air is flowing, the pressure reading at the airway includes two forces stacked on top of each other: the force required to overcome resistance in the tubing, endotracheal tube, and airways, plus the force required to stretch the lung tissue and chest wall. Once the ventilator briefly holds the breath at its peak, airflow stops and the resistance component drops away. The pressure that remains, measured during this “end-inspiratory hold,” is the plateau pressure. It is expressed in centimeters of water (cm H₂O).

The distinction between peak pressure and plateau pressure matters because only the stretching component is thought to drive the kind of tissue damage that worsens outcomes. In patients with high airway resistance, peak pressure can be dramatically higher than plateau pressure. A study of infants with severe respiratory viral infections found that peak inspiratory pressure overestimated plateau pressure by an average of about 7 cm H₂O, because the small, inflamed airways created enormous resistance that vanished once flow stopped.¹ In adults with conditions like severe asthma or chronic obstructive lung disease, the gap can be even wider. Treating peak pressure as if it were plateau pressure leads to unnecessary alarm or, worse, misguided ventilator adjustments.

How the Measurement Is Performed

The standard technique is straightforward in concept: at the end of a machine-delivered breath, the ventilator closes both the inspiratory and expiratory valves for a short pause, typically half a second to a few seconds. With no air entering or leaving, the pressure in the circuit equilibrates and settles to a relatively flat line on the monitor. That flat reading is recorded as plateau pressure.

How long you hold the pause matters. In patients with acute respiratory distress syndrome (ARDS) breathing without any added baseline pressure, plateau pressure measured at half a second was about 20 cm H₂O and drifted down to roughly 18 cm H₂O by five seconds, a small but statistically meaningful decline caused by slow redistribution of air among lung regions with different mechanical properties.² In practice, most ICU protocols use a hold of about half a second to two seconds. Shorter holds risk catching the tail end of flow; longer holds can be uncomfortable for awake patients and may slightly understate the pressure relevant to fast-moving lung injury.

An alternative approach uses mathematical modeling of the expiratory flow curve to estimate plateau pressure continuously without pausing the ventilator at all. One method based on the expiratory time constant performed comparably to the traditional hold maneuver across various ventilator modes, with the advantage of providing a real-time, automatic readout rather than requiring a clinician to manually trigger each measurement.³ This kind of automation could make continuous plateau-pressure surveillance practical, rather than relying on spot checks a few times a day.

Why Keeping It Low Prevents Lung Damage

Ventilator-induced lung injury, or VILI, is one of the central hazards of mechanical ventilation. It happens through several overlapping mechanisms: overstretching of lung tissue, repeated collapse and reopening of small air sacs, and the release of inflammatory signals that can spill into the bloodstream and damage distant organs.⁴ Plateau pressure became a clinical target because it was the most accessible bedside number thought to correlate with how much the lung tissue was being stretched.

The commonly cited threshold is 30 cm H₂O. A meta-regression analysis looking at ARDS patients during the first week in the ICU found that short-term mortality was significantly higher when plateau pressures exceeded 32 cm H₂O during the first three days of admission. Pressures below 27 or 30 cm H₂O were not independently associated with a reduction in death risk, suggesting that the benefit of keeping pressure low has a floor: below a certain point, further reductions do not measurably help.⁵ Interestingly, the association between plateau pressure and mortality was strongest on day one and weakened by day seven, hinting that early ventilator settings carry outsized importance.

Some clinicians have pushed the target even lower in specific situations. A randomized trial in elderly patients undergoing open abdominal surgery tested a strategy of adjusting tidal volume to keep plateau pressure at or below 20 cm H₂O, combined with moderate positive end-expiratory pressure. The approach improved ventilation parameters and postoperative recovery compared to a conventional tidal volume strategy.⁶ Whether such aggressive targets should be routine outside of surgical anesthesia remains debated, but the trial illustrates a broader trend: tailoring ventilator settings to pressure feedback rather than rigid volume prescriptions.

Driving Pressure and Why It May Matter More

Plateau pressure tells you the total elastic recoil pressure of the respiratory system at end-inspiration, but it does not account for the baseline pressure already present at the start of each breath. Most ventilated patients receive positive end-expiratory pressure (PEEP), a small amount of pressure held in the lungs between breaths to prevent collapse. Subtracting PEEP from plateau pressure gives the driving pressure, which reflects how much additional stretch each breath imposes on the lung above its resting state.

A landmark analysis of over 3,500 ARDS patients from nine randomized trials found that driving pressure was the ventilation variable most strongly associated with survival. Each increment of about 7 cm H₂O in driving pressure was linked to a roughly 40 percent increase in the relative risk of death. The striking finding was that this association held even in patients whose plateau pressures and tidal volumes were already within the “protective” range.⁷ In other words, a patient could have a plateau pressure of 28 cm H₂O, well under the traditional 30 cm H₂O ceiling, and still face elevated risk if most of that pressure was driving pressure rather than PEEP.

This finding reshaped how many intensivists think about ventilator targets. Plateau pressure remains relevant, but treating it as the sole safety metric can miss patients who are being harmed by excessive tidal stretch relative to their available lung volume. One pilot study measuring compliance by different methods found that the standard technique using the inspiratory hold may actually overestimate compliance and underestimate driving pressure, meaning some patients could be receiving more harmful ventilation than the numbers suggest.⁸ The field has not settled on a universal driving pressure cutoff, but values above roughly 13 to 15 cm H₂O increasingly raise concern.

When Plateau Pressure Misleads

Plateau pressure reflects the combined elastic properties of the lungs and the chest wall. In a thin patient with a compliant chest wall, nearly all of the measured pressure is transmitted to the lung tissue, so plateau pressure is a reasonable stand-in for lung stress. But in a patient whose chest wall is stiff, whether from obesity, abdominal distension, or chest wall edema, a large fraction of the measured pressure is “used up” pushing against the chest wall and never reaches the lung at all.

This has real clinical consequences. A study of morbidly obese patients on mechanical ventilation found that plateau pressure significantly overestimated the actual transpulmonary pressure, the force truly distending the lung. Neither body mass index nor intra-abdominal pressure reliably predicted how much of the plateau pressure was attributable to the chest wall versus the lung.⁹ The implication is that following a strict 30 cm H₂O plateau-pressure limit in these patients could lead clinicians to accept dangerously low tidal volumes or inadequate PEEP, when the lung itself is nowhere near being over-stretched.

Research measuring both airway and esophageal pressures in patients with acute respiratory failure confirmed how variable chest wall contribution can be. The estimated transpulmonary pressure during an inspiratory hold averaged about 18 cm H₂O, but the spread was enormous, with a standard deviation over 10 cm H₂O, meaning some patients had very low lung stress and others had dangerously high stress at identical plateau pressure readings.¹⁰ The authors concluded that airway pressure alone may not adequately predict how much the lung is being distended.

A similar limitation shows up in a different framing: lung stress and strain. In ARDS patients, a given plateau pressure produced widely variable transpulmonary pressures depending on the ratio of lung elastance to total respiratory system elastance, which ranged from 0.33 to 0.95 across patients.¹¹ At one extreme, two-thirds of the airway pressure was absorbed by the chest wall; at the other, virtually all of it reached the lung. The researchers went so far as to conclude that plateau pressure and tidal volume are “inadequate surrogates for lung stress and strain.”

Measuring What Actually Reaches the Lung

The most direct way to estimate true lung stress at the bedside is esophageal manometry. A thin balloon catheter placed in the esophagus measures pressure that approximates pleural pressure, the pressure surrounding the lung. Subtracting this value from the airway pressure gives the transpulmonary pressure, which represents the actual force stretching the lung tissue. The technique is imperfect because esophageal pressure is influenced by the weight of mediastinal structures and by patient positioning, but it adds a dimension of information that airway pressure alone cannot provide.

A pilot trial tested a ventilator strategy in which PEEP was set to maintain a positive end-expiratory transpulmonary pressure, rather than relying on standard airway pressure targets alone. Compared to conventional management, the esophageal-pressure-guided group showed significantly better oxygenation and lung compliance.¹² The results were encouraging enough to spark larger trials, though adoption in everyday practice remains limited. Placing and maintaining an esophageal balloon adds complexity, cost, and a learning curve that not every ICU team has embraced.

Pediatric data tells a parallel story. In children with ARDS, changes in airway pressure correlated poorly with changes in transpulmonary pressure, and assuming a “safe” airway pressure ceiling of 30 cm H₂O produced transpulmonary pressures that ranged from a genuinely safe zone into potentially harmful territory depending on the individual child’s chest wall mechanics.¹³ Animal data, by contrast, showed a cleaner linear relationship between plateau pressure and lung stress in both healthy and injured lungs, possibly because experimental animals tend to have more uniform chest wall properties than real patients.¹⁴

The Problem With Spontaneous Breathing Efforts

The inspiratory hold maneuver assumes the patient is passive, letting the ventilator do all the work. When patients are breathing spontaneously on top of ventilator-delivered breaths, their own muscle activity contaminates the measurement. A contracting diaphragm generates negative pleural pressure that pulls against the positive pressure in the airway, making the plateau reading unreliable.

Recent work has tried to quantify how much spontaneous effort distorts the number. In a bench-and-patient study, the time needed to reach a stable plateau increased dramatically with higher inspiratory effort: under passive conditions, plateau pressure stabilized within about half a second, but during strong spontaneous effort it could take more than 1,500 milliseconds.¹⁵ Holding the breath for that long in a spontaneously breathing patient is often impractical and uncomfortable. A multicenter analysis confirmed that muscular activity frequently makes inspiratory holds unreadable during patient-triggered breaths, and that unreadable plateaus were associated with higher levels of inspiratory effort.¹⁶

This creates a practical gap: many ICU patients are on assisted modes of ventilation where they initiate and contribute to each breath. In those patients, the very measurement that clinicians rely on for safety is unreliable. The workaround in some units is brief sedation boluses to achieve passive conditions before measuring, but this is disruptive and may not reflect the pressures the patient experiences during normal minute-to-minute ventilation.

Pediatric Considerations

Children on mechanical ventilation are often managed with pressure-targeted modes rather than volume-targeted modes, partly because pediatric endotracheal tubes are small and uncuffed, making volume delivery less predictable. In pressure-targeted ventilation, the flow decelerates as the set pressure is reached, and the displayed peak pressure is sometimes treated as a rough substitute for plateau pressure. ARDS guidelines for children suggest limiting plateau pressure to about 28 to 30 cm H₂O, but because plateau pressure is most accurately measured in constant-flow (volume control) modes, it is rarely obtained in pediatric practice.¹⁷ Instead, clinicians rely on peak inspiratory pressure limits as a proxy, which introduces the same estimation errors seen in infants with high airway resistance.

How Prone Positioning Helps

Turning an ARDS patient face-down, known as prone positioning, has become one of the strongest evidence-based interventions for severe cases. Part of the benefit is mechanical: in the supine position, the weight of the heart and abdominal organs compresses the dependent (back) portions of the lung, while the front portions may become overdistended. Flipping the patient redistributes these forces, producing a more even gas-to-tissue ratio throughout the lung and a more homogeneous distribution of stress and strain. The mortality reduction seen with proning is attributed partly to less overdistension in the nondependent regions and less cyclical collapse in the dependent regions.¹⁸

Physiological studies have confirmed the mechanical effects directly. Compared to supine ventilation, prone positioning reduced peak and plateau transpulmonary pressures, decreased lung elastance, increased the resting lung volume by roughly a quarter, and reduced the strain each breath imposed on the tissue.¹⁹ In practical terms, the same ventilator settings produce less harmful mechanical force on the lung when the patient is prone, which partly explains why proning can improve outcomes even without changing the numbers on the ventilator screen.

Ultra-Protective Ventilation With Extracorporeal Support

If the goal is to minimize lung stretch, the logical extreme is to reduce ventilator support to an absolute minimum and let an external machine handle gas exchange. Veno-venous extracorporeal membrane oxygenation (VV-ECMO) does exactly this: blood is drawn out of the body, oxygenated and cleared of carbon dioxide by a membrane, and returned to the venous circulation. With ECMO handling the metabolic work, the ventilator can be dialed down to settings that would otherwise leave the patient dangerously under-ventilated.

Two randomized controlled studies in severe ARDS patients on VV-ECMO tested this “ultra-lung-protective” strategy, combining very low tidal volumes, low respiratory rates, and consequently very low plateau pressures and driving pressures. Both found that the ultra-protective approach significantly reduced markers of pulmonary inflammation, a process known as biotrauma, which is thought to be a key driver of ongoing lung injury and multi-organ failure.²⁰ ²¹ Whether reduced biotrauma translates into better survival remains an open question, but the biological logic is compelling: less mechanical force means less inflammatory signaling, which means less secondary damage.

Diagnostic Artifacts and False Readings

Even the act of measuring can introduce errors. One well-documented artifact is “pseudo auto-PEEP,” where an end-expiratory occlusion maneuver suggests the presence of trapped air and elevated baseline pressure, but independent measurements of airway pressure do not confirm it. A case series described this discrepancy in critically ill patients and flagged it as a potential source of misdiagnosis.²² If a clinician acts on a falsely elevated auto-PEEP reading, they might inappropriately adjust ventilator settings or fluid management, treating a measurement artifact rather than a real physiological problem.

Other sources of error include leaks around the endotracheal tube cuff, which prevent a true equilibrium during the hold; cardiac oscillations that create small pressure swings during the pause, making it hard to identify the true plateau; and the timing issues discussed earlier when the patient is not fully passive. Experienced respiratory therapists learn to read around these artifacts, but in busy units with high patient-to-nurse ratios, an erroneous plateau reading can easily go unquestioned and influence treatment for hours before the next check.